A control optimization method considering economic and comfort constraints

Abstract

This invention discloses a control optimization method considering economic and comfort constraints, comprising: real-time acquisition of environmental data inside and outside the classroom and operating power, acquisition of environmental data from weather forecasts, and calculation of the data error between the environmental data from weather forecasts and the real-time acquired environmental data; if the data error does not exceed an error threshold, the power control data pre-optimized under the conditions of energy consumption economic objectives and indoor comfort constraints using an improved multi-objective gray wolf algorithm is directly selected as the current control parameter; if the data error exceeds the error threshold, a correction amount is calculated based on the real-time acquired environmental data and operating power, and the pre-optimized power control data is corrected by the correction amount and used as the current control parameter.

Description

Technical Field

[0001] This invention relates to the field of split-type air conditioner control optimization technology, and in particular to a control optimization method that takes into account economic and comfort constraints. Background Technology

[0002] Split-type air conditioners (mostly inverter compressors + indoor units) are typically distributed across multiple rooms in classrooms, offices, and similar settings. Users set the temperature (e.g., 26℃), mode (cooling / dehumidifying), and fan speed via remote control. Internally, the air conditioner uses manufacturer strategies (similar to constant temperature + hysteresis / PI, etc.) to adjust the compressor frequency, ensuring the return air temperature is close to the set value. In school classrooms, energy saving is generally achieved by manually turning on the unit at set times, pre-cooling it, and turning it off after class.

[0003] However, the key to energy saving in split-type air conditioners often lies in the continuous scheduling of compressor power / frequency. Relying solely on setting the temperature and starting / stopping the system can easily lead to fluctuations of "overcooling - then rising again - then forced cooling again," resulting in high energy consumption and unstable comfort.

[0004] Furthermore, traditional temperature control focuses primarily on temperature, neglecting the impact of humidity, wind speed, radiation, and clothing / metabolism on comfort. In classrooms with varying population density and strong solar radiation, the same temperature setting may correspond to different levels of comfort. Even when comfort modes are available, they are often based on manufacturer experience curves and do not constrain quantifiable comfort indicators such as thermal sensation among individuals. These experience-based rules typically either prioritize energy saving, resulting in unstable comfort, or prioritize comfort, leading to higher energy consumption.

[0005] Furthermore, in a school classroom setting, there are many disturbance factors, such as people entering and exiting, opening and closing doors and windows, changes in outdoor temperature, solar radiation, and equipment heat dissipation. The thermal response of split-type air conditioners and rooms is time-varying, and the performance of fixed proportional-integral-derivative (PID) control or fixed rules fluctuates greatly under different weather conditions / number of people, which may lead to comfort exceeding limits or increased energy consumption.

[0006] The purpose of this invention is to design a control optimization method that considers economic and comfort constraints to address the problems existing in the prior art. Summary of the Invention

[0007] In view of this, the purpose of this invention is to propose a control optimization method that takes into account economic and comfort constraints, and can solve the above-mentioned problems.

[0008] This invention provides a control optimization method considering economic and comfort constraints, comprising:

[0009] S1 collects real-time environmental data inside and outside the classroom, operating power, obtains environmental data from weather forecasts, and calculates the data error between the environmental data from weather forecasts and the real-time collected environmental data.

[0010] S2 If the data error does not exceed the error threshold, the power control data obtained in advance through the improved multi-objective gray wolf algorithm under the constraints of energy consumption economy and indoor comfort is directly selected as the current control parameter;

[0011] If the data error exceeds the error threshold, S3 calculates the correction amount based on the real-time collected environmental data and operating power, corrects the pre-optimized power control data using the correction amount, and uses it as the current control parameter.

[0012] The beneficial effects of this invention are:

[0013] First, by constructing a multi-objective optimization model for energy consumption and comfort and setting hard constraints with the comfort index (PMV), the problem of traditional air conditioning control being unable to simultaneously take into account energy saving and comfort, and the difficulty in quantifying and guaranteeing comfort, is solved, thereby achieving low-energy operation within the range of PMV constraints.

[0014] Secondly, by improving the multi-objective gray wolf algorithm for offline optimization and combining it with the approximation of the ideal solution to select the comprehensive optimal solution from the Pareto front, the problem that conventional single-objective or empirical parameter tuning strategies are prone to getting trapped in local optima and are difficult to obtain the optimal control parameters that balance economy and comfort is solved, thus achieving efficient optimization of the power control sequence and output of the optimal solution.

[0015] Third, by introducing an adaptive infeasible solution repair mechanism based on the degree of comfort exceedance, the problem of candidate solutions easily violating constraints under strong comfort constraints, resulting in a low proportion of feasible solutions and a decrease in optimization efficiency is solved, thereby improving the feasible solution generation rate and the quality of optimization results.

[0016] Fourth, by setting forecast and measured error thresholds and adopting an online power correction strategy, the problem of reduced control effectiveness of offline optimization strategies due to weather forecast data deviations is solved, achieving dynamic correction and robust control of power parameters without compromising comfort constraints. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings required in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the method in Example 1.

[0019] Figure 2 This is a flowchart of step S2 in Example 1.

[0020] Figure 3This is a comparison chart of different iterative solution results for a sunny day (Tmax < 35℃) in Example 1.

[0021] Figure 4 This is a comparison chart of different iterative solution results for the high temperature (Tmax>35℃) in Example 1.

[0022] Figure 5 This is a comparison chart of the different iterative solution results for rainy days in Example 1.

[0023] Figure 6 This is a circuit diagram of the microprocessor and anti-interference capacitor element in Embodiment 2.

[0024] Figure 7 This is a circuit diagram of the protection circuit in Embodiment 2. Detailed Implementation

[0025] To facilitate understanding by those skilled in the art, the structure of the present invention will now be described in further detail with reference to the accompanying drawings. It should be understood that, unless otherwise specified, the order of the steps mentioned in this embodiment can be adjusted according to actual needs, and they can even be executed simultaneously or partially simultaneously.

[0026] Example 1

[0027] like Figure 1 As shown, this embodiment of the invention provides a control optimization method considering economic and comfort constraints, including:

[0028] S1 collects real-time environmental data inside and outside the classroom, operating power, obtains environmental data from weather forecasts, and calculates the data error between the environmental data from weather forecasts and the real-time collected environmental data.

[0029] The S101 collects real-time temperature data and solar radiation data inside and outside the classroom, as well as the air conditioning operating power.

[0030] S102 obtains temperature data and solar radiation data at the same time and location from weather forecasts, and calculates the temperature data error and solar radiation data error.

[0031] In this step, the number of people in the classroom and the weather are key factors affecting the control of air conditioning use and comfort. Outdoor temperature directly determines the heat transfer driving force of the building envelope (walls, windows, infiltration, etc.). Predicted temperatures that are too high or too low will cause deviations in the indoor temperature changes calculated by the equivalent thermal parameter model, leading to insufficient or excessive cooling of the power sequence originally generated based on the forecast, affecting comfort and energy consumption. Solar radiation is a significant heat gain term entering the room through windows, especially in classrooms with sunlight, causing rapid and nonlinear changes in heat gain. Deviations in radiation forecasts will directly cause errors in the solar heat gain term in the model, resulting in a large deviation in the estimation of the required cooling capacity of the air conditioning.

[0032] S2 If the data error does not exceed the error threshold, the power control data obtained in advance through the improved multi-objective gray wolf algorithm under the constraints of energy consumption economy and indoor comfort is directly selected as the current control parameter;

[0033] In this step, since the control strategy is predetermined and the data source is weather forecast values, there is a certain degree of error. To ensure that the temperature changes in the classroom match the user's actual needs, the compressor's power control needs to be adjusted. To reduce the number of ineffective power adjustments, a threshold is set for the error between the forecast and actual values, and the pre-analyzed power control data is selected based on the actual situation.

[0034] like Figure 2 As shown, the power control data, pre-optimized using an improved multi-objective gray wolf algorithm under the constraints of energy consumption economy and indoor comfort, is obtained through the following steps:

[0035] S201 calculates the decision dimension, and initializes multiple candidate power sequences based on the decision dimension and the Tent mixed mapping. ;

[0036] S2011 controls the time interval of the split air conditioner's power. and the total duration of the course Calculate decision dimensions The calculation formula is as follows:

[0037] ;

[0038] In this step, the time interval for controlling the power of the split-type air conditioner is... It can be 1 minute. Since the optimization object is a power sequence over a period of time, rather than a single power value, each power value corresponds to a control time slice. The total course duration was divided into Only after a period of time did we realize how many decision variables (i.e., the power sequence length) needed to be optimized, which directly determined the dimension and initialization length of the gray wolf position vector.

[0039] S2012 sets the iteration count, population size, and archive size of the gray wolf population, initializes the hunting vectors A and C, and initializes multiple candidate power sequences using Tent mixed mapping. .

[0040] In this step, in the position update (hunting / surveying) of the Grey Wolf algorithm, the power sequence position of an individual (wolf) is represented by a vector. The hunting vector A is used to control the stride and direction between exploration and development, and the hunting vector C is used to introduce random weights to perturb the hunting distance of the leader wolf and enhance diversity. A and C are needed to complete the first position update in the 0th iteration. At the same time, they are usually random vectors that are "re-randomized in each iteration and in each dimension". Initialization is to give the algorithm an executable starting point and connect it with the subsequent dynamic update rules.

[0041] Conventional multi-objective gray wolf algorithms typically initialize wolf pack positions (power sequences) randomly, which may result in uneven distribution and insufficient early exploration. Using Tent mixed-variable mapping to initialize gray wolf positions makes the initial solutions more evenly distributed within the feasible region, improving diversity and global search capabilities, thus leading to a faster and easier attainment of the optimal solution.

[0042] S202 calculates the energy consumption target and comfort target for each candidate power in each candidate power sequence;

[0043] The overall energy consumption of all split-type air conditioners within the S2021 statistical control period is used as the energy consumption target for the corresponding candidate power, and the calculation formula is as follows:

[0044] ,

[0045] in, , , These are the time point at which the split-type air conditioner was calculated, the room number it is located in, and the number of air conditioners in the room. , , These are the total duration of control of the split-type air conditioner, the total number of rooms, and the total number of air conditioners in each room. For public buildings at time nodes First The first room The power consumption of the air conditioner To determine the time step for solving, This represents an optimization operator, indicating the search for optimal candidate power control schemes under satisfied constraints, such that... Take the minimum;

[0046] In this step, the economic objective is... The main statistical method is to measure the overall energy consumption of split-type air conditioners during the control period, thereby evaluating the energy-saving effect of the control strategy.

[0047] S2022 inputs the candidate power into the equivalent thermal parameter model to calculate the corresponding temperature data. Based on the temperature data, it calculates the predicted average thermal perception index (PMV) of the current environment as the comfort target. The calculation formula is as follows:

[0048] ,

[0049] in, M Metabolic rate, unit is In summer, 1.2 Met is often used in classrooms. Mechanical power, unit is , This is the surface area coefficient for clothing; in classrooms, 0.5clo is commonly used for summer clothing. Air temperature, in °C. It is the partial pressure of water vapor. The mean radiant temperature is expressed in °C. The surface temperature of the garment is expressed in degrees Celsius (°C). The convective heat transfer coefficient;

[0050] ,

[0051] in, j , k These represent the time point and room number of the split-type air conditioner being calculated. J , K These refer to the total duration of control of the split-type air conditioner and the total number of rooms, respectively. To determine the time step for solving, The optimization operator represents the search for the optimal control scheme among all candidate control schemes. The minimum solution, where T represents the effective statistical duration.

[0052] In this step, the ETP (Equivalent Thermal Parameters) model is a simplified thermodynamic model that approximates the heat transfer process of a room / building using a small number of "equivalent" thermal resistance-heat capacity networks. It treats the room as consisting of several heat storage bodies (heat capacity) and heat transfer channels (thermal resistance). By converting power into temperature through the equivalent thermal parameter model, the predicted average thermal perception index PMV can be further calculated.

[0053] The Predicted Mean Thermal Sensation (PMV) index was used to represent the comfort evaluation index for each classroom. Although individual perceptions of comfort vary, the PMV is a comprehensive evaluation model established by Fanger after calculating the basic equation of human thermal balance and combining it with the differences in psychological and physiological thermal comfort among different people. Essentially, it represents the predicted results of a group of people voting on comfort in a particular environment. Because the PMV considers multiple factors related to human thermal comfort, it can basically meet the comfort requirements of most people.

[0054] Based on this, the national standard GB / T18049-2017 also adopts the Predicted Average Thermal Sensation (PMV) index to analyze and measure thermal comfort. According to the relevant data in GB / T18049-2017, combined with the air temperature and relative humidity collected during classroom use, the PMV index of the classroom can be calculated. The smaller the absolute value of the PMV index, the better the comfort.

[0055] objective function To calculate the average predicted thermal perception index (PMV) after the air conditioner has been running for 20 minutes, in order to prevent the poor comfort level when the air conditioner is first turned on from interfering with the accurate assessment of room comfort.

[0056] S203 If the predicted comfort target of the candidate power exceeds the constraint, adaptive infeasibility solution repair is achieved by adjusting the candidate power at the previous time step. The calculation formula is as follows:

[0057] ,

[0058] in, and Let represent the power of the infeasible solution before and after repair at node t, respectively. This refers to the rated power of the air conditioner. This represents the predicted mean thermal perception index (PMV) of the comfort target at time t+1, based on the current candidate power sequence. Indicates the adjustment factor;

[0059] In this step, the constraint is: |PMV|≤0.5. The operating frequency of an air conditioner compressor is generally between 10% and 100%. Operating below 10% of the cooling power can easily damage the compressor. Therefore, when calculating the air conditioning power for each room, the following condition must be met:

[0060] ,

[0061] in, To be at the time node No. The operating power of the air conditioner in each classroom and Let the minimum and maximum operating power of the air conditioner in the k-th room be 10% and 100% of the rated power of the air conditioner, respectively.

[0062] Considering the inherent differences in perceived thermal sensations among individuals, the Predicted Average Thermal Sensitivity (PMV) index cannot directly and uniformly represent the feelings of everyone. Therefore, Fanger conducted further statistical analysis. When the PMV index is greater than +1 or less than -1, people in the current environment will experience discomfort; conversely, when the PMV index falls between +1 and -1, people in the current environment will feel comfortable. Therefore, Fanger proposed a Predicted Per-centage Dissatisfaction (PPD) index to represent people's level of dissatisfaction with the thermal environment. Its expression is as follows:

[0063] ,

[0064] The PPD (Predicted Average Thermal Perceived Value) index shows that when the predicted average thermal perception value (PMV) is approximately -0.5 or +0.5, a PPD of 10% indicates that about 10% of people feel dissatisfied with the current environment. Even when the predicted average thermal perception value (PMV) is the most comfortable level of 0, the PPD is still approximately 5%, meaning that about 5% of people still feel dissatisfied. This index illustrates that while air conditioning can hardly achieve complete thermal comfort for everyone, it can ensure that the majority of people in a given area feel comfortable. Therefore, the national standard GB / T18049-2017 uses PPD ≤ 10% as the basis for evaluating air-conditioned environments; that is, a thermal environment where more than 90% of people feel satisfied is considered a thermally comfortable environment.

[0065] According to the national standard requirement of PPD ≤ 10%, the predicted average thermal perception index (PMV) falls within the range of [-0.5, 0.5]. Therefore, except at the initial moment of turning on the air conditioner, the predicted average thermal perception index (PMV) calculated for the indoor environment should meet the following constraints:

[0066] ,

[0067] in, To be at the time node No. The comfort level within each classroom.

[0068] To optimize the control of split-type air conditioners in classrooms, an improved multi-objective gray wolf algorithm was used to optimize the multi-objective optimization model. Although the constraints in the optimization solution are relatively few, the formula... A strong constraint, which is difficult to implement, is imposed: if the predicted average thermal perception index (PMV) at any time point does not meet the constraint, the entire control scheme of the gray wolf will be considered infeasible, thus rendering the gray wolf's search invalid. Therefore, it is necessary to perform a feasibility-making repair on all infeasible solutions.

[0069] The infeasibility arises because the air conditioning power was too low at the previous time node, causing the predicted average thermal perception index (PMV) at that moment to not meet the actual requirements. Therefore, an adaptive infeasibility repair method based on the predicted average thermal perception index (PMV) is adopted. Depending on whether the predicted average thermal perception index (PMV) at this time node exceeds the constraints, the power of the previous time node is adjusted to attempt to repair individual infeasible time nodes in the infeasibility solution, minimizing the number of invalid searches by the gray wolf.

[0070] S204 If the comfort target of the candidate power does not exceed the constraints, then the solutions are non-dominated based on the Pareto dominance relation, the non-dominated solutions are archived, and three representative non-dominated solutions are selected from the archive as leader wolves according to the roulette wheel method. ;

[0071] In this step, diversity metrics (such as crowding distance / grid density) for each non-dominated solution in the archive are calculated, and the probability of selection based on these metrics is assigned accordingly. Roulette wheel selection is a probability-based sampling method where the probability of each candidate solution being selected is proportional to its "weight." A higher weight makes it easier to be selected. In the Grey Wolf algorithm's archive selection, roulette wheel selection is often used to favor solutions in sparser regions to maintain diversity. For example, crowding distance is used as a weight; a larger crowding distance (located in a "hollower" region at the Pareto front) increases the probability of being selected as the leader wolf, thus preventing all leader wolves from crowding into the same small segment at the front.

[0072] Because multi-objective optimization typically lacks a single "best" solution, energy consumption and comfort often conflict. The role of non-dominated sorting is to establish a hierarchy of excellence across multiple objectives: solutions are divided into F1, F2, ..., where F1 represents solutions in the current population / archive that are not dominated by any other solution (the optimal front), F2 represents solutions dominated only by F1, and so on. The leader wolf comes from the set of non-dominated solutions, thus guiding the search towards the Pareto front. Without non-dominated sorting, it's difficult to define "better" in the conflict between "energy consumption vs. comfort," and the algorithm easily degenerates into a single-objective search biased towards one particular goal.

[0073] S205 selects the leader wolf from the non-dominated solution archive. The candidate power sequence of each gray wolf in the current population is updated according to the gray wolf optimal position update formula to obtain the updated population power sequence.

[0074] S2051 If the current iteration count reaches the reverse learning iteration count, then for the leader wolf... Generate and evaluate the reverse solution; if the reverse solution is better than the original leader wolf... Then update the leader wolf using the reverse solution. And update the archive;

[0075] S2052 If the current iteration number has not reached the reverse learning iteration number, check the maximum iteration number. If the maximum iteration number has not been reached, proceed to the next iteration.

[0076] In this step, backward learning is a strategy to enhance global exploration. For the current solution, we simultaneously consider its "opposite point" in the search space. That is, when the current optimal solution is near a certain region, we also look at whether there is a better one in the "opposite region", which helps to escape local optima.

[0077] Because reverse learning incurs additional computational overhead (generating new solutions, handling constraints, and simulating / computing objectives), performing it in every generation is costly and may disrupt convergence. Therefore, reverse learning is triggered based on the number of iterations. Furthermore, to ensure the algorithm terminates within a finite time / computing capacity and avoid infinite loops, the maximum number of iterations is used as the termination condition.

[0078] If the number of iterations reaches the maximum number of iterations, S206 generates a Pareto front and uses an approximation of the ideal solution to select the optimal power allocation scheme.

[0079] S2061 performs standardized decision matrix processing on all candidate power Pareto front solutions, calculated as follows:

[0080] ,

[0081] in, For the elements of the decision matrix, For the first The solution is at the th solution. The elements of the original decision matrix represent the objectives. and The respective The maximum and minimum values ​​of each target;

[0082] In this step, the optimal power allocation scheme is selected from the Pareto front using the approximation ideal solution TOPSIS. The principle is to calculate the relative proximity of each Pareto front solution to the ideal solution and the negative ideal solution, and then rank them to select the relatively optimal solution. First, all Pareto front solutions are processed using a standardized decision matrix. The purpose of this is to address the incomparability between different dimensions, allowing for comparisons of different objectives on the same scale.

[0083] S2062 calculates the Euclidean distance between the Pareto front solution of each candidate power and its ideal solution and its negative ideal solution, using the following formula:

[0084] ,

[0085] ,

[0086] in, and The first The distance between each solution and the positive and negative ideal solutions. and The first The positive and negative ideal solutions for each objective;

[0087] In this step, the ideal solution is found. and negative ideal solution These are the distance benchmarks used to measure each Pareto front solution relative to the ideal solutions. The positive ideal solution is the minimum of all objectives, i.e., minimum energy consumption and minimum predicted mean thermal index (PMV); the negative ideal solution is the maximum of all objectives, i.e., maximum energy consumption and maximum predicted mean thermal index (PMV). The Euclidean distances between each Pareto front solution and the ideal and negative ideal solutions are calculated.

[0088] S2063 selects the relative proximity by calculating the relative proximity of the Pareto front solution for each candidate power. The solution with the largest value is taken as the optimal solution, which is the final air conditioning control strategy. The calculation formula is as follows:

[0089] ,

[0090] in, For the first The relative closeness of a solution to the ideal solution is such that a larger value indicates that the solution is closer to the ideal solution.

[0091] In this step, the technique for ordering preference by similarity to the ideal solution is a multi-criteria decision-making method: among multiple alternatives, the solution closest to the "ideal optimal solution" and furthest from the "negative ideal worst solution" is selected. In multi-objective optimization (e.g., "minimum energy consumption" and "optimal comfort"), the output at the end of the iteration is typically a Pareto front set of non-dominated solutions, rather than a single optimal solution. There is no "absolute optimality" in the Pareto set; some solutions have lower energy consumption but slightly lower comfort, while others offer better comfort but higher energy consumption. Engineering control must be implemented with a clear strategy; air conditioning control ultimately outputs an executable power sequence / control command.

[0092] Table 1 Energy consumption results for three weather conditions

[0093]

[0094] Table 2. Predicted Average Thermal Sensation Index (PMV) Results for Three Types of Weather

[0095]

[0096]

[0097] According to the results in Tables 1 to 3, the energy consumption of air conditioners has decreased under the improved algorithm optimization control. However, the rate of energy consumption reduction varies depending on the weather. This is because the comfort target is different under different weather conditions, which leads to different target temperatures for energy saving. In particular, the energy consumption of air conditioners is low in rainy weather and high in hot weather, so the energy-saving space is not as large as in sunny weather.

[0098] Since this control strategy is applied in public buildings such as schools, which have a large number of rooms and various types of air conditioners, certain requirements are placed on the algorithm's solution speed and quality. Three algorithms (the improved multi-objective gray wolf algorithm, the multi-objective gray wolf algorithm, and the NSGA-II algorithm) were trained with different numbers of iterations for three different weather conditions. The results are as follows: Figures 3 to 5 As shown.

[0099] By comparison Figures 3 to 5 It can be concluded that the improved multi-objective gray wolf algorithm and the multi-objective gray wolf algorithm of this application can find the local frontier solution by relying on gray wolf encirclement when the number of iterations is relatively small. However, as the number of iterations increases, the optimization ability of the multi-objective gray wolf algorithm is slightly weaker than that of the NSGA-II algorithm. However, the improved multi-objective gray wolf algorithm of this paper has a similar optimization ability to the NSGA-II algorithm after the number of iterations increases, and even surpasses the NSGA-II algorithm in some application scenarios. Therefore, the improved multi-objective gray wolf algorithm of this paper is suitable for the formulation and solution of this multi-objective strategy.

[0100] If the data error exceeds the error threshold, S3 calculates the correction amount based on the real-time collected environmental data and operating power, corrects the pre-optimized power control data using the correction amount, and uses it as the current control parameter.

[0101] S301 calculates the changes in heat load and radiative heat gain based on the errors in temperature data and solar radiation data, respectively.

[0102] S302 calculates and converts the power compensation amount based on changes in heat load and radiative heat gain. The power compensation amount is then superimposed on the power control data obtained in advance through the improved multi-objective gray wolf algorithm. The calculation formula is as follows:

[0103] ,

[0104] in, and These represent the power before and after the adjustment, For the number of walls, For the first A wall; and These represent the actual outdoor temperature and the forecast outdoor temperature, respectively. and These represent the actual solar radiation intensity and the predicted solar radiation intensity, respectively. and For the first The equivalent outdoor resistance and area of ​​the wall.

[0105] In this step, the control device corrects the power by comparing the difference between the measured and predicted outdoor temperature and solar radiation, and offsets the excess heat generated indoors by increasing the air conditioning power.

[0106] The energy consumption and comfort levels of the two rooms were statistically analyzed, as shown in Table 4. The improved algorithm in this paper can dynamically adjust the control strategy under different weather conditions to achieve a balance between energy consumption and comfort, with energy savings reaching 20% ​​at certain time points.

[0107] Table 4 Energy Consumption and Comfort Indicators

[0108]

[0109] Example 2

[0110] This embodiment provides a control optimization system that considers economic and comfort constraints, including:

[0111] The sensor unit, whose output is electrically connected to the main control unit, is used to collect environmental data.

[0112] Furthermore, the sensor unit includes:

[0113] Indoor and outdoor temperature sensors are used to acquire indoor and outdoor temperature data;

[0114] A solar radiation intensity sensor is used to acquire solar radiation data.

[0115] The operating current detection sensor is used to obtain the operating current of the split air conditioner, and then to monitor the operating power.

[0116] In this embodiment, the sensor unit primarily monitors temperature, solar radiation, and power. Specifically, the sensor unit collects data such as indoor and outdoor temperature, solar radiation intensity, and compressor power to verify and adjust the control strategy.

[0117] The main control unit is used to generate power control parameters based on the environmental data acquired by the sensor unit and send the values ​​to the actuator unit.

[0118] Furthermore, the main control unit includes: a microprocessor and an anti-interference capacitor element;

[0119] The power input terminal of the microprocessor is connected to several anti-interference capacitors to suppress interference conducted through the power lines.

[0120] Furthermore, a protection circuit is provided between the sensor unit and the main control unit;

[0121] The protection circuit includes: a ceramic gas venting pipe, a self-resetting fuse, and a transient voltage suppressor connected in sequence to the output terminal of the sensor unit;

[0122] The ceramic gas venting pipe is used to release abnormal transient overvoltages;

[0123] The self-resetting fuse is used to provide decoupling and ensure reliable operation of the ceramic gas venting pipe;

[0124] The transient voltage suppressor is used to quickly clamp residual overvoltage and protect the input circuit of the main control unit.

[0125] Several actuator units are electrically connected to the main control unit to execute the power control parameters issued by the main control unit.

[0126] In this embodiment, the actuator unit is a frequency converter based on a three-phase inverter bridge circuit. Currently, most variable frequency air conditioners on the market use brushless DC motors, which are permanent magnet excitations. Compared with AC asynchronous motors, they have advantages such as small size, simple structure, and high efficiency. Therefore, a three-phase inverter bridge circuit is used to perform frequency conversion and power control of the brushless DC motor.

[0127] To prevent high-frequency interference from the frequency converter from significantly affecting the circuitry of the main control unit and other supporting components, anti-interference measures are implemented from both circuit design and device assembly perspectives.

[0128] The STM32F103RCT6TR microprocessor is the core of the main control unit. To avoid coupling with the inverter and other equipment, maintain stable power supply voltage output, and reduce the influence of external factors on the main control unit, capacitors C18, C20, C22, and C23 connected to the power supply are added for decoupling. The principle is to utilize the low impedance characteristics of capacitors for high-frequency signals to provide a bypass path for noise, preventing interference from propagating through the power lines. Figure 6 As shown.

[0129] The circuit used in conjunction with the sensor unit is as follows: Figure 7 As shown. Signal transmission between all units utilizes the universal RS485 communication standard, primarily used in industrial control and network transmission. Since the air conditioner compressor is located outdoors, a 3R090A-5S ceramic gas vent pipe is used at the front end to discharge abnormal transient overvoltages. A JK-nSMD010 / 60V self-resetting fuse is used in the middle for decoupling, making the front ceramic gas vent pipe more easily activated. An SMBJ15CA transient voltage suppressor is used at the rear end to protect downstream circuits from damage by abnormal overvoltages.

[0130] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0131] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0132] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0133] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0134] It should be noted that any reference signs placed between parentheses in the claims should not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The words first, second, and third, etc., do not indicate any order. These words can be interpreted as names.

[0135] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0136] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0137] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0138] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms should not be construed as necessarily referring to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

Claims

1. A control optimization method considering economic and comfort constraints, characterized in that, A control optimization system that considers both economic and comfort constraints includes: The sensor unit, whose output is electrically connected to the main control unit, is used to collect environmental data. The main control unit is used to generate power control parameters based on the environmental data acquired by the sensor unit and send them to the actuator unit. Several actuator units are electrically connected to the main control unit to execute the power control parameters issued by the main control unit; The method includes: S1 collects real-time environmental data inside and outside the classroom, operating power, obtains environmental data from weather forecasts, and calculates the data error between the environmental data from weather forecasts and the real-time collected environmental data. S2 If the data error does not exceed the error threshold, the power control data obtained in advance through the improved multi-objective gray wolf algorithm under the constraints of energy consumption economy and indoor comfort is directly selected as the current control parameters, including: S201 calculates the decision dimension and initializes multiple candidate power sequences based on the decision dimension and the Tent chaotic mapping; S202 calculates the energy consumption target and comfort target for each candidate power in each candidate power sequence; S203 If the predicted comfort target of the candidate power exceeds the constraint, adaptive infeasibility solution repair is achieved by adjusting the candidate power at the previous time step. The calculation formula is as follows: ， in, and Let represent the power of the infeasible solution before and after repair at node t, respectively. This refers to the rated power of the air conditioner. This represents the predicted mean thermal perception index (PMV) of the comfort target at time t+1, based on the current candidate power sequence. Indicates the adjustment factor; S204 If the comfort target of the candidate power does not exceed the constraints, then the solutions are non-dominated based on the Pareto dominance relation, the non-dominated solutions are archived, and three representative non-dominated solutions are selected from the archive as leader wolves according to the roulette wheel method. ; S205 selects the leader wolf from the non-dominated solution archive. The candidate power sequence of each gray wolf in the current population is updated according to the gray wolf optimal position update formula to obtain the updated population power sequence. S206 If the number of iterations reaches the maximum number of iterations, then the Pareto front solution is generated, and the optimal power allocation scheme is selected by approximating the ideal solution. If the data error exceeds the error threshold, S3 calculates the correction amount based on the real-time collected environmental data and operating power, corrects the pre-optimized power control data using the correction amount, and uses it as the current control parameter.

2. The control optimization method considering economic and comfort constraints according to claim 1, characterized in that, The process of collecting real-time environmental data inside and outside the classroom, operating power, obtaining environmental data from weather forecasts, and calculating the data error between the environmental data from weather forecasts and the real-time collected environmental data includes: The S101 collects real-time temperature data and solar radiation data inside and outside the classroom, as well as the air conditioning operating power. S102 obtains temperature data and solar radiation data at the same time and location from weather forecasts, and calculates the temperature data error and solar radiation data error.

3. The control optimization method considering economic and comfort constraints according to claim 1, characterized in that, The calculation of the decision dimension, and the initialization of multiple candidate power sequences based on the decision dimension and the Tent chaotic mapping, include: S2011 calculates the decision dimension by controlling the time interval of the split air conditioner power and the total duration of the course; S2012 sets the number of iterations, population size, and archive size of the gray wolf population, initializes the hunting vectors A and C, and uses Tent chaotic mapping to initialize multiple candidate power sequences.

4. The control optimization method considering economic and comfort constraints according to claim 1, characterized in that, The calculation of the energy consumption target and comfort target for each candidate power in each candidate power sequence includes: The overall energy consumption of all split-type air conditioners within the S2021 statistical control period is used as the energy consumption target for the corresponding candidate power, and the calculation formula is as follows: ， in, , , These are the time point at which the split-type air conditioner was calculated, the room number it is located in, and the number of air conditioners in the room. , , These are the total duration of control of the split-type air conditioner, the total number of rooms, and the total number of air conditioners in each room. For public buildings at time nodes First The first room The power consumption of the air conditioner To determine the time step for solving, This represents an optimization operator, indicating the search for optimal candidate power control schemes under satisfied constraints, such that... Take the minimum; S2022 inputs the candidate power into the equivalent thermal parameter model to calculate the corresponding temperature data. Based on the temperature data, it calculates the predicted average thermal perception index (PMV) of the current environment as the comfort target. The calculation formula is as follows: ， in, M Metabolic rate, unit is In summer, 1.2 Met is often used in classrooms. Mechanical power, unit is , This is the surface area coefficient for clothing; in classrooms, 0.5clo is commonly used for summer clothing. Air temperature, in °C. It is the partial pressure of water vapor. The mean radiant temperature is expressed in °C. The surface temperature of the garment is expressed in degrees Celsius (°C). The convective heat transfer coefficient; , in, j , k These represent the time point and room number of the split-type air conditioner being calculated. J , K These refer to the total duration of control of the split-type air conditioner and the total number of rooms, respectively. To determine the time step for solving, The optimization operator represents the search for the optimal control scheme among all candidate control schemes. The minimum solution, where T represents the effective statistical duration.

5. The control optimization method considering economic and comfort constraints according to claim 1, characterized in that, The leader wolf selected from the non-dominated archive. The candidate power sequence of each gray wolf in the current population is updated according to the gray wolf optimal position update formula, and the updated population power sequence includes: S2051 If the current iteration number reaches the reverse learning iteration number, then for the leader wolf... Generate and evaluate the reverse solution; if the reverse solution is better than the original leader wolf... Then update the leader wolf using the reverse solution. And update the archive; S2052 If the current iteration number has not reached the reverse learning iteration number, check the maximum iteration number. If the maximum iteration number has not been reached, proceed to the next iteration.

6. The control optimization method considering economic and comfort constraints according to claim 1, characterized in that, If the number of iterations reaches the maximum number of iterations, a Pareto front solution is generated, and the optimal power allocation scheme is selected using an approximation of the ideal solution, including: S2061 performs standardized decision matrix processing on all candidate power Pareto front solutions, calculated as follows: ， in, For the elements of the decision matrix, For the first The solution is at the th solution. The elements of the original decision matrix represent the objectives. and The respective The maximum and minimum values ​​of each target; S2062 calculates the Euclidean distance between the Pareto front solution of each candidate power and its ideal solution and its negative ideal solution, using the following formula: ， ， in, and The first The distance between each solution and the positive and negative ideal solutions. and The first The positive and negative ideal solutions for each objective; S2063 calculates the relative proximity of the Pareto front solution for each candidate power and selects the solution with the highest relative proximity as the optimal solution, which is the final air conditioning control strategy. The calculation formula is as follows: ， in, For the first The relative closeness of a solution to the ideal solution is such that a larger value indicates that the solution is closer to the ideal solution.

7. The control optimization method considering economic and comfort constraints according to claim 1, characterized in that, The step of calculating correction amounts based on real-time collected environmental data and operating power, correcting the pre-optimized power control data using these correction amounts, and then using the corrected data as the current control parameters includes: S301 calculates the changes in heat load and radiative heat gain based on the temperature data error and the solar radiation data error, respectively. S302 calculates and converts the changes in heat load and radiative heat gain into a power compensation amount, which is then superimposed onto the power control data obtained in advance through the improved multi-objective gray wolf algorithm. The calculation formula is as follows: ， in, and These represent the power before and after the adjustment, For the number of walls, For the first A wall, and These represent the actual outdoor temperature and the forecast outdoor temperature, respectively. and These represent the actual solar radiation intensity and the predicted solar radiation intensity, respectively. and For the first The equivalent outdoor resistance and area of ​​the wall.

8. The control optimization method considering economic and comfort constraints according to claim 1, characterized in that, The main control unit includes: a microprocessor and an anti-interference capacitor element; The power input terminal of the microprocessor is connected to several anti-interference capacitors to suppress interference conducted through the power lines.

9. The control optimization method considering economic and comfort constraints according to claim 8, characterized in that, A protection circuit is provided between the sensor unit and the main control unit; The protection circuit includes: a ceramic gas venting pipe, a self-resetting fuse, and a transient voltage suppressor connected in sequence to the output terminal of the sensor unit; The ceramic gas venting pipe is used to release abnormal transient overvoltages; The self-resetting fuse is used to provide decoupling and ensure reliable operation of the ceramic gas venting pipe; The transient voltage suppressor is used to quickly clamp residual overvoltage and protect the input circuit of the main control unit.

Images